A Behavioral Finance Approach to Student Academic Performance: The Role of Screen Time and Sleep.

Abstract

This study examines how behavioral factors impact students' academic performance. Using Structural Equation Modeling (SEM), Bayesian regression, and Ordinary Least Squares (OLS), the research explores both direct and indirect effects of daily time allocation on exam outcomes. The analysis focuses on variables such as study hours, social media use, Netflix viewing, sleep duration, and constructs of productive versus leisure use of time. SEM results show that mental health (β = 0.355, p < 0.001), exercise frequency (β = 0.167, p < 0.001), and sleep hours (β = 0.133, p < 0.001) significantly enhance productive time use, which strongly predicts higher exam scores (β = 1.114, p < 0.001). Leisure activities like social media and Netflix have a moderate negative effect on academic performance (β = -0.395, p = 0.002). Bayesian regression confirms these findings, with study hours showing a strong positive effect (mean = 12.663, HDI = [12.090, 13.244]) and social media (mean = -2.786) and Netflix hours (mean = -2.229) exhibiting strong negative impacts. Sleep hours have a moderate positive effect (mean = 2.219).The OLS model explains 71.4% of the variance in exam scores, identifying study hours/day (β = 0.432, p < 0.001) and sleep (β = 0.188, p < 0.001) as significant positive predictors, while social media (β = -0.071) and Netflix hours (β = -0.046) contribute negatively. These findings reflect behavioral finance principles, where time and cognitive resources are allocated under bounded rationality, emphasizing the cost of leisure screen time on academic outcomes.